DSSAT

DSSAT v4.5 - Canegro Sugarcane Plant Module

User Documentation

M. Jones1 and A. Singels2

South African Sugarcane Research Institute

Mount Edgecombe, South Africa

[email protected]

[email protected]

International Consortium for Sugarcane Modelling (ICSM)

http://sasri.sasa.org.za/misc/icsm.html

December 2008

2

Published by:

South African Sugarcane Research Institute

170 Flanders Drive, Mount Edgecombe, 4300

Private Bag X02, Mount Edgecombe, 4300

Tel: (031) 508 7400 Fax: (031) 508 7597

Website: www.sugar.org.za/sasri

© December 2008

First printed: 2008

Copyright subsists in this work. No part of this work may be reproduced in any form or by any means

without the publisher‟s written permission. Once written approval has been sought, this publication may

only be reproduced in its entirety and no sections may be removed. Whilst every effort has been made

to ensure that the information published in this work is accurate, SASRI takes no responsibility for any

loss or damage suffered by any person as a result of the reliance upon the information contained

therein.

http://www.sugar.org.za/sasri

3

Contents

1. Introduction ......................................................................................................................................... 4

2. Inputs .................................................................................................................................................. 4

2.1. Soil ......................................................................................................................................... 5

2.2. Weather .................................................................................................................................. 8

2.3. Management ........................................................................................................................ 12

2.4. Cultivar ................................................................................................................................. 32

2.5. Experimental data ................................................................................................................ 47

3. Simulation settings ............................................................................................................................ 50

3.1. Specific Sugarcane simulation options ................................................................................ 50

3.2. Irrigation ............................................................................................................................... 50

4. Running the model and viewing outputs ........................................................................................... 51

4.1. Running the sugarcane model ............................................................................................. 51

4.2. Viewing and graphing model output..................................................................................... 52

4.3. General discussion of model outputs ................................................................................... 55

5. Simulating a sequence of plant and ratoon crops and a fallow period. ............................................ 55

5.1. Introduction .......................................................................................................................... 55

5.2. Scenario ............................................................................................................................... 56

5.3. Method ................................................................................................................................. 56

6. Acknowledgements ........................................................................................................................... 57

7. References ........................................................................................................................................ 57

4

1. Introduction

This document guides the user to set up and execute a Canegro simulation run within the DSSAT

environment, manipulate weather soil and crop management input data, calibrate plant input

parameters, and compare simulated output data with observed data. It takes the form of a tutorial,

where a complete simulation is set up from scratch. Each step in this process is described. Emphasis is

placed on sugarcane-specific aspects of the DSSAT system; further comprehensive documentation,

covering general aspects of operating the DSSAT software, is provided with the DSSAT distribution.

Every DSSAT simulation consists of an Experiment file (FileX), which references soil (FileS) and

weather (FileW) files. This document describes how to go about creating a FileX for running sugarcane

simulations, as well as providing some information on creating weather and soil files, particularly where

these have special relevance for sugarcane.

The Canegro model in DSSAT makes use of genetic information defined in species, ecotype and

cultivar files. Some guidance for creating new cultivar and ecotype definitions is presented in this

document.

2. Inputs

Every DSSAT simulation consists of an Experiment file (FileX), which defines crop management for a

particular „experiment‟ (set of model runs or „treatments‟) and references soil (FileS) and weather

(FileW) files. These resources are separated in this way because soil definitions and weather data can

be used in several simulations (even for different crops), whereas the experimental setup file is unique

to a particular experiment. This document describes how to go about creating a FileX for running

sugarcane simulations, as well as providing some information on creating weather and soil files,

particularly where these have special relevance for sugarcane.

Crop modelling has a complex nature. While DSSAT is among the most user-friendly systems available

for crop modelling, care must be taken to ensure that the simulations are reliably set up. Several

resources potentially need to be defined – soil, weather, crop management, cultivar/genetic information

and simulation options. Technical errors can creep in within any of these resources. Scientific errors are

not considered (i.e. the scientific bases for the model are assumed to be correct). DSSAT can,

however, assist in detecting technical input errors, but this is beyond the scope of this document.

These technical errors can be minimised, and more importantly, more easily detected, if an iterative

process is followed when setting up a simulation. This process includes the following:

Define the management first – associate the treatments within an experiment with existing soil and

weather data, existing cultivars, and standard simulation options. See if the model runs and check

for any errors. Use of these „placeholders‟ is recommended because they have been tested and shown

in general to be free of technical errors.

Step 1: Define the weather data; update the management file (FileX) such that these new weather

data are used instead of the „placeholder‟ weather data set up in step 1. Test to ensure that the

simulation still runs and results make sense.

Step 2: Set up any special simulation options (e.g. specifying FAO-56 PET method). Run the

simulation to check that it still works and makes sense.

Step 3: Define the soil; update the management file (FileX) such that this new soil profile definition(s)

is/are used instead of the „placeholder‟ soil data set up in step 1. Test to ensure that the simulation still

runs and results make sense.

5

Step 4: Set up new cultivars by copying an existing one and updating the variables one-by-one; update

the management file (FileX) such that these new cultivar definitions are used instead of the

„placeholder‟ cultivars set up in step 1. Test to ensure that the simulation still runs and results make

sense.

Step 5: The simulation setup is now complete.

If the model ceases to operate at any step in this process, the user will know immediately where the

error has occurred and remedial action can be taken without delay.

If such an iterative process is not followed, the user runs the risk of not knowing where an error has

occurred – the model might simply „crash‟, leaving the user with the potentially large task of tracking

down the error(s).

An alternative error-minimising approach is to first create each of the resources – soil, weather,

cultivars, etc. – and then modify a known functioning simulation to use these. Again, an iterative

approach is used – one change is made, the model is run, results checked, and then the next change is

made. If, for example, the model ran fine with one soil and then crashed on the new soil profile, it could

be inferred that the new soil definition has an error and should be checked.

2.1. Soil

The DSSAT SBuild program is used for creating soil files. Click on the Sbuild icon on the main DSSAT

screen to load this software.

Each soil in DSSAT is defined as a soil profile and stored in a soil file. Soils are named and coded. The

soil code is used in the experiment file to refer to soil information. Each soil profile has a number of soil

layers. Each layer is associated with specific physical and chemical characteristics. DSSAT-Canegro

only uses physical aspects. The physical water-holding effects of Soil Organic Matter are modelled by

the DSSAT v4.5 system and might affect Canegro simulations.

The most important soil variables are:

2.1.1. Water holding characteristics

DUL – drained upper limit. This is the maximum stable soil water content the soil can maintain. It is

equivalent to Field Capacity. It is determined by watering the soil very well, covering it and then leaving

it for several days, after which the water content is determined gravimetrically (or otherwise).

LL – lower limit. This is soil water content of the soil at which the plant can no longer extract water from

the soil. It is equivalent to the Permanent Wilting Point. It can be determined using a pressure plate

system, and is soil water content at which no more water can be extracted at 15 bars of pressure.

SAT – saturated water capacity. This is the highest soil water content above which water will

immediately drain from the profile.

Units: all soil water holding measures are volumetric, i.e. volume of water per volume of soil, typically

cm3/cm3 .

2.1.2. Infiltration characteristics

SWCN – saturated water conductivity. This is the speed with which water traverses from one soil layer

to a lower one. In a wet soil, surface water infiltration rate is dependent on the SWCN of the slowest

layer (cm/hr).

6

2.1.3. Rooting characteristics

WR – root distribution weighting. Root distribution is plant and soil dependent since root mass and

volume in sugarcane tend to decline with depth. If the soil has an impermeable layer or a water table,

roots will not grow below that layer.

In addition to these, soil albedo (light reflectivity) and Curve Number (for runoff calculations) need to be

set.

DSSAT prompts for entry of nutritional and soil organic matter data, but DSSAT-Canegro does not

model nutrients. If you have this information, it is worth entering it, because future model updates will

hopefully include nutrient support, and this soil definition can also be used for other crops that do model

nutrients.

DSSAT also stores other soil information that may not be used by the model – e.g. soil colour. Again, it

is good practice to enter all known soil information because this is a useful database of soil information

and these variables may be used in future.

There are several optional parameters, some of which are used only as of v4.5 of the DSSAT software

(e.g. the effect of soil organic matter on moisture retention of the soil).

2.1.4. Creating a new soil

Step1. Open Sbuild.

Step 2. Access the Profile menu and select „New‟.

Step 3. Enter general soil data. This is used for selecting the soil and for various calculations (e.g.

runoff).

7

Step 4. Create all layers first (by clicking on the „Add Layer‟ button several times) – then characterise

them, e.g. if there are 10 soil layers, add all 10 first and then enter values.

Step 5. Look at the Exercise handout – you will see the number of layers (7).

Step 6. Click Next.

Step 7. Then define soil layer thicknesses (depths of lower layer boundaries); Note: start from the

deepest layer and work your way up.

Step 8. Click Next.

Step 9. Then enter the attribute values – e.g. DUL, LL, etc (remember the soil albedo etc. at the top!).

Step 10. Click on „Finish‟.

8

Step 11. VERY IMPORTANT! – click on Profile Save. If you forget to do this, you will lose the

information you have entered

Step 12. Then click „File Save‟.

Step 13. If you look in C:\DSSAT4\Soil, in the Soil.sol file, you should see your new soil.

Notes:

i. Always remember to open/save PROFILES, which are within soil FILES.

ii. To work with a different FILE, use the File menu.

iii. You may have to copy the updated soil.sol file into the sugarcane directory if the model has trouble

finding it.

iv. If you cannot find the new soil profile in XBuild, click on the „Refresh‟ menu in Xbuild.

v. To work with a different PROFILE, use the profile menu.

vi. The „quirky‟ nature of SBuild may make it necessary to modify the file directly (e.g. using Notepad):

Ensure that your columns line up correctly.

Ensure that the changes / additions you make to the files are as consistent as possible with

the existing information in the files.

Files are fixed-format, so exact numbers of spaces are important!

Use spaces, not the TAB key.

2.2. Weather

Use the WeatherMan software to enter weather data – click on the WeatherMan icon on the main

DSSAT screen to load the program.

Variables required for good sugarcane modelling are:

Bare minimum:

Minimum temperature (oC)

Maximum temperature (oC)

Rainfall (mm)

Solar Radiation (MJ/m2)

9

The following are highly recommended:

Maximum relative humidity (%)

Windspeed (km/day)

or

Dewpoint temperature (oC)

The last one/two variables are for the FAO-56 Potential Evapotranspiration calculation. If you do not

have these variables, you will have to use the less accurate Priestley-Taylor method (an Xbuild setting

– see the management section for this).

WeatherMan makes provision for only one measure of humidity; we decided that that would represent

maximum humidity, and is assumed to occur at the same time as minimum temperature. This allows the

model to calculate Dewpoint Temperature. If dewpoint temperature is calculated more accurately

outside of DSSAT (e.g. with wet and dry bulb temperatures), this should be imported into WeatherMan,

because it will be more accurate than the internal calculation.

The easiest way to enter information into WeatherMan is to import a text file of weather data:

Step 1. Import the text file into Excel.

Step 2. Assuming the data are in a spreadsheet, select the appropriate columns of the weather data

and paste into a blank Excel worksheet.

10

Step 3. Then copy these columns and paste into a Notepad window. Save this file as a text file, e.g.

„Station25_data.txt‟.

Step 4. Use the WeatherMan wizard to import this data into WeatherMan:

Step 5. Click on „New station‟:

Step 6. Select the „Input or import raw weather data and save as a new station‟ option.

11

Step 7. Then click File Open.

Step 8. Browse to the text file you have just created, select it, click ok.

Step 9. The screen will look something like this:

Step 10. You will be presented with your data in grid form, but it will need to be manipulated slightly

first:

Define the columns (right-click on the column header and select a variable and units).

When all columns are defined, remove the header row.

12

Note: In the example above, an Excel formula was used to create the YRDOY date format. This is not

strictly necessary, but ensures that WeatherMan is not confused by the computer’s regional date

format. For this reason, it is recommended to try to reformat dates into this format. In general, it is

recommended that data are entered in a form as close as possible to the final stored form – and any

processing required should be performed outside of WeatherMan. Rainfall should be in mm/day,

temperatures in degrees C, solar radiation in MJ/m2, humidity in percentages; column headings can

also be correctly named (DATE, RAIN, SRAD, RHUM, WIND, TMIN, TMAX).

2.3. Management

Setup of management information is very important and perhaps also the most difficult. This difficulty is

associated with actually using the program (which occasionally displays unexpected behaviour) and

conceptual difficulties associated with factor levels and treatments. The technical difficulties can be

mitigated somewhat by carefully and closely following these instructions.

2.3.1. Concepts of factors, levels and treatments

An agricultural experiment is usually intended to compare different crop management approaches, with

the intention of e.g. deriving descriptive information to advise growers, or perhaps inferring information

about the physiology of the plant. Each „management approach‟ is a treatment – a combination of

different management factor levels. For example, if an agronomist is interested in the effect of different

rowspacings and cultivars of sugarcane on growth and development, he will be assessing two

management factors (rowspacing and cultivars). If he compares three different rowspacings (e.g. 0.6,

1.0 and 1.4 m) and two cultivars (e.g. NCo376 and R570), he will have three factor levels within the

rowspacing management factor and two factor levels within the cultivar factor. He will have six

treatments – the six unique combinations of the management factor levels:

Table 1. Management factor row spacing

Level Value

1 0.6 m

2 1.0 m

3 1.4 m

13

Table 2. Management factor cultivar

Level Value

1 NCo376

2 R 570

Table 3. Treatments

Treatment Level Row spacing factor level

Cultivar factor level

Description

1 1 1 0.6 m rowspacing, cultivar NCo376

2 1 2 0.6 m rowspacing, cultivar R570





In DSSAT, all treatments are constructed in this way. It is first necessary to define each management

factor level. Each level is assigned a level number, and the user can associate a short description to

each level to aid identification when it comes to constructing the treatments. Defining the treatments –

as combinations of management factor levels – is done last.

Note: Rather than providing every possible management factor for treatments, DSSAT arranges

management factors into categories. Row spacing, for example, falls into the „planting‟ management

factor. If several row spacings are to be assessed, rather than have different rowspacing factor levels,

the user will have different planting factor levels; each planting factor level will be characterised by

different row spacings.

2.3.2. Setting up an experiment (management) file in DSSAT using XBuild

The following steps describe how an experiment – the definition of treatments with factor levels

(management) – is set up. The Xbuild program is used; a FileX is created.

Use of Xbuild is described via an example. The details of this simulation are as follows:

Location: Mount Edgecombe, South Africa

Years: 1989, 1990

Row spacing: 1.2 m

Cultivar NCo376

Harvest at 12 months

Treatments: irrigated and non-irrigated, April and October ratooning

14

4 treatments:

Level Plant date Irrigation/rainfed

1 Treatment 1 – start 1 April 1989 rainfed

2 Treatment 2 – start 1 October 1989 rainfed

3 Treatment 3 – start 1 April 1989 irrigated

4 Treatment 4 – start 1 October 1989 irrigated

Step 1. Open DSSAT and create a new experiment.

Step 2. Xbuild should load. Enter the following information:

15

i. Choose „Experimental‟ for File Type

ii. Enter a short description of the experiment

iii. Enter a 2-letter institute code

iv. Enter a 2-letter site code

v. Enter a meaningful year; either the simulated year of the experiment or the year that the simulation is being run

vi. Enter an experiment number; choose something appropriate

vii. Very importantly, choose „Sugarcane‟ as the Crop.

16

viii. Enter general information (follow on-screen indications); ignore the other sections of the screen.

ix. Click Next

x. Click on FileSave. Using the information you have entered, DSSAT will assign a unique code

for this experiment. This code will be used as the filename for this FileX.

Remember, Xbuild caters for ALL crops. For this reason, it must support all of the inputs required for all

crops. Not all of these inputs are used by the sugarcane model. If you have this information, however, it

is worth entering, because it is possible that a future version of the sugarcane model will use that input.

Unfortunately, despite certain inputs not being used by the sugarcane model, Xbuild itself requires

certain inputs to be set. In such cases, a default value should be entered (some examples will follow).

Inputs that are not used by the sugarcane model cannot be used as factor levels, because the model

will effectively run exactly the same simulation for every such level.

17

Step 2. Choose cultivars to use in the simulation:

Click on the Add button to add another cultivar factor level. Click on the cultivar code in the rightmost

column of the Cultivar table to get a drop-down to select which cultivar you want to use. In this example

simulation, NCo376 was used.

Step 3. Field setup.

Note: if you are not presented with the Fields screen, click on the ‘Environment’ menu and then click on

‘Fields’.

The field has two basic components – a soil, and a location.

i. Choosing a soil

Choose a soil from the drop-down list on the right-hand side of the screen. If you have not yet created a

soil profile/file for this experiment, simply choose any soil from the list. It is a placeholder until you have

set up the correct soil, after which you will come back to this screen and choose the correct soil instead.

Select a surface texture if you know this.

The other soil parameters can be ignored, but if you have the information it is worth entering it.

Remember, although this is basically just a model configuration file you are setting up, it can contain

enough information that it becomes a really useful tool for storing historical experiment/trial information.

ii. Choosing a location (weather station)

The location is defined by choice of weather station: as it is the weather that is driving the model, it is

assumed that the weather station is at the trial site. In practice, weather stations tend to be some way

away from the trial site, but this is unavoidable. If you have not yet entered weather data for this

experiment, either choose the next nearest station available on the drop-down list, or choose any

station. For now, this is a placeholder; when you have entered weather data for this experiment, you will

come back and update this so that the correct weather data is used.

18

You will need to choose an 8-character FieldID. At the top right of the screen, you will see that you are

currently editing Level 1 of the Fields management factor. You can (and should) enter a short

description of this field. If your experiment treatments require more than one weather station or soil, you

will need to add new field factor level(s) using the Add button.

19

Click OK.

Step 4. Set up the planting details.

Note 1: if you are not presented with the Planting Details screen, click on the ‘Management’ menu and

then click on ‘Planting’.

Note 2: Xbuild has a bug in it where it misreads dates unless they are entered in the American date

format. Xbuild definitely works best if you have the Short Date Format of your PC set to the US

convention of MM/DD/YYYY. This can be set in the Windows Control Panel – Start MenuControl

PanelRegional & Language Settings (icon)Customise (button)Date (tab)Short date (drop-down

list)

In this example, planting date is one of the factors varied for the treatments. Two factor levels are

required – planting on 1 April 1989 and 1 October 1989.

i. Enter a description for this planting factor level.

ii. Type in the planting date in the planting date textbox (see Note 2 above). You will see the available

weather data years for the currently-selected station listed on the left-hand side.

iii. Omit emergence date.

iv. Select „Dry seed‟ if this is a sugarcane plant crop (no „sett‟ option exists) or „Ratoon‟ if this is a

ratoon crop.

v. Select „Rows‟ from planting distribution.

20

vi. It is necessary to enter a seed population, despite this not being used by the model; choose any

value (preferably the correct value).

vii. Ignore population at emergence and row direction.

viii. Enter row spacing (in cm).

ix. Enter Planting Depth (in cm); again, this is not used by the model but DSSAT will crash if it does

not find this input. 20 cm is a plausible default.

x. Then add the new factor level for the October planting:

21

xi. Make the appropriate changes such that it correctly reflects what we are trying to achieve with the

treatments; (1) give the new level a meaning short description, (2) add the new planting date and

(3) click OK to complete:

You have now created a „bare bones‟ simulation. Xbuild will return to its main screen. If you have not

yet saved, this is a very good time to do so: FileSave.

Further work is required to define a basic practically-functional simulation:

Harvest dates must be set for sugarcane. Failure to do so will result in the model running until the

weather data runs out.

Note: Unlike many other crops, sugarcane, in most climatic circumstances, does not „mature‟ and die,

so the farmer must choose when to harvest. In contrast, Maize, for example, will calculate its own

harvest date and DSSAT will inform the user when the crop was harvested.

The treatments need to be defined as combinations of factor levels.

Step 6. Set up harvest dates

It is now necessary to set harvest dates. This can be achieved in two ways - by date or by age at

harvest.

i. Click on the „Management‟ menu and select „Harvest‟:

22

ii. Give the harvest factor level a name/description (e.g. „April crop harvest‟)

iii. Click on the drop-down menu and select „On reported date(s)‟, even if it already says this (this is an

Xbuild quirk).

Note: if you wanted to specify a number of days after planting to harvest, select the Age/Days from

planting option from this menu.

iv. Enter the date of harvest in the Date column. The date format issue applies here too – input MUST

be in the USA-style MM/DD/YYYY format.

v. Press OK and Save.

23

If you are using more than one Harvest factor level:

Now repeat step (i) to re-access the Harvests screen

Check that the date was entered correctly for factor level 1

Click the Add button at the above mid-left to add a new Harvest factor level

Enter a description (e.g. „October harvest‟)

Repeat steps (iii – v)

24

The harvest factor levels have now been set up. Defining the treatments is all what remains of setting

up the basic simulation.

Step 7. Defining treatments

In the example, four treatments need to be set up. So far, the following have been defined:

Two planting factor levels (for the two planting dates)

One field factor level

One cultivar factor level

Two harvest factor levels (corresponding with the planting dates)

This is only enough for two treatments – the other two are for irrigation, which needs to be set up. As

this is a simulation option, it will be ignored for the time being. The user is invited to go ahead and

define the two rainfed treatments prior to setting up irrigation.

Procedure for defining the two treatments:

i. Click on the Treatments menu.

ii. Type in a name for the first treatment (level 1).

iii. Click on the table cell in the cultivar column, treatment 1 row; a drop-down of cutivars should

appear. Select one.

iv. Click on the cell in the Field column. Select the field you have just defined.

v. Click on the cell in the Plant column and select the first planting date.

25

vi. In the „Harv.‟ Column cell, select the first harvest level.

vii. Click OK, and the FileSave.

viii. Click on the Treatments menu again.

ix. Click on the little arrow on the leftmost extreme of the table row defining Treatment 1.This will

highlight the treatment.

x. Click on the Add button. This will create a copy of the first treatment, but numbered „2‟.

xi. Repeat step (v) but select the second planting date.

xii. Repeat step (vi), but select the second harvest date.

xiii. Click OK and save.

Step 8. Defining simulation options and irrigation

Defining irrigation in Xbuild sits awkwardly between crop management and DSSAT simulation controls.

Please read the section on Simulation Options before proceeding. Other simulation options MUST be

set, even if no irrigation is planned.

26

First, general simulation options must be set:

i. Click on the Simulation Options menu.

ii. Set the simulation start date to the earliest planting date; click OK and save.

iii. Re-open Simulation Options (step i).

iv. Rename DEFAULT SIMULATION OPTIONS to „Rainfed simulation options‟.

v. Click OK and save, then reopen Simulation Options (step i).

vi. Set planting to „On reported date‟ in the Management section (button):

27

vii. Tell DSSAT to harvest on reported dates; click on the Management button, then on the harvest

button, and select “On reported date(s)” from the drop-down menu.

viii. Click OK and save.

ix. Re-open Simulation Options (step i).

x. Tell DSSAT to simulate water but not nutrients (Canegro does not yet support nutrients – all

simulations are presumed to have adequate fertiliser): Click on the Options button, in the water

drop-down select „Yes‟, and make sure everything else is set to „No‟.

xi. Click OK and save.

xii. Re-open Simulation Options (step i).

xiii. Click on the Methods button.

xiv. Select WeatherMeasured data.

xv. In the Evapotranspiration drop-down, select FAO-56. This is important for sugarcane, so long

as humidity and windspeed or dewpoint data are available in the weather dataset. If not, select

Priestley-Taylor. Click OK, Save, reopen, etc.

Now, the irrigation options can be set.

i. Add a new Simulation Options level and rename to „irrigated simulation options‟. This will be a

copy of the rainfed options.

ii. Click OK, save, and reopen, and select the Irrigated simulation option factor level by clicking on

the arrow on the left of the factor levels list.

iii. Click on the „Management‟ button.

iv. Click on the „Irrigation and water management‟ button.

28

v. Select „Automatic when required‟, and appropriate values for the management depths and

thresholds.

vi. Click OK, save.

vii. Click on the treatments menu.

viii. For the two existing treatments, select the „Rainfed simulation option‟ factor level in the „Sim

Contr.‟ Column:

ix. Select the first treatment and click Add.

x. Rename this to „April planting, irrigated‟; modify the „Sim. Contr.‟ for this to reference the

Irrigated simulation control.

xi. Select the second treatment, click Add, rename to „October planting, irrigated‟; modify the „Sim.

Contr.‟ for this to reference the Irrigated simulation control.

xii. Check that the set up factor levels for each treatment make sense.

29

xiii. Click OK and Save.

Setup of this example experiment is now complete. The experiment should run. If the model runs and

produces some output, all is well. Check the output to ensure that the planting/harvest dates operated

as intended and that the correct number of treatments ran, etc. Soil definition and weather files may

now need to be created and this experiment file modified to reference these new data.

Step 9. Updating FileX to use new soil profile

Having set up this experiment and established that it is functioning, and then having set up the soil

profile following the instructions in section 2.1, it is now necessary to update the FileX (experiment) to

use the new soil.

i. Assuming you are no longer in Xbuild (save and close if necessary), load up the main DSSAT

program (if it is not already open).

ii. Navigate to the sugarcane directory using the navigation tree on the mid-left of the screen.

iii. Click on the refresh button to ensure that the experiment you have just set up is displayed:

iv. You will now see a complete list of sugarcane simulations in the upper right pane of the DSSAT

window.

v. Locate the experiment (by code or description).

vi. Click on the checkbox next to it; the treatments you have set up should appear in the pane

below:

30

vii. Click on the name of the simulation (the row should then be highlighted).

viii. Right-click and select „Edit file‟ (the file will open in XBuild) [If you select „View file‟, it will open

in Notepad]:

ix. Refresh Xbuild by clicking on the Refresh menu in the Xbuild program:

31

x. Click on the „Environment‟ menu, „Fields‟.

xi. Select the new soil from the drop-down list (if the soil is not listed, try Refreshing again (step

ix)).

xii. Then click on FileSave, and then close Xbuild.

xiii. Refresh the experiment list in DSSAT (step iii).

xiv. If the experiment is rerun, it should use the new soil and give different results.

Step 10. Updating FileX to use new weather data

Updating the FileX to use new / correct weather data is very similar to updating for a new soil.

i. Follow steps i x above.

ii. In the weather station list, select the station just set up (if it does not appear, try Refreshing

Xbuild (step 9, iii) again.

iii. File Save.

iv. Close Xbuild.

v. Re-run the simulation – results should be different.

32

Step 11. Updating the FileX to use a new / different cultivar

If a new cultivar definition is set up, Xbuild needs to be updated in a similar manner as soil and weather

files.

i. Follow steps i x under Step 9.

ii. Access the cultivar screen (Management Cultivars).

iii. Select the new cultivar, OK, File Save

Having thus set up the experiment file (FileX), and the soil and weather files (and possibly the cultivar

file), and then updated the FileX to use these, the experiment setup is now complete.

2.4. Cultivar

2.4.1. Introduction

Genetic parameters are used to capture the genetic control of how sugarcane plants respond to

environmental and management factors. These are normally grouped into three categories, namely

species (identical values for all cultivars), ecotype (identical values for groups of similar cultivars) and

cultivar parameters (specific to cultivars).

With the DSSAT Canegro project we attempted to transfer all implicit parameters in the code to the

appropriate parameter files. The genetic parameterization of sugarcane is new ground, and insufficient

information and knowledge is available to have necessarily rationally allocated all parameters to the

species, ecotype and cultivar categories. The approach was to allocate the majority of parameters to

the ecotype and cultivar categories to allow user access. Although this resulted in a large number of

parameters in these two categories, we believe that the 22 parameters contained in the cultivar file

provides sufficient flexibility to assess the impact of these on crop growth and development. The 26

parameters in the ecotype file provide further flexibility if needed, although these parameters are

sometimes quite empirical and difficult to determine.

Twenty-three parameters were allocated to the species category based on circumstantial evidence (see

Table 3.1 in scientific documentation). As more cultivar information and knowledge about the

mechanisms of genotypic control of crop response to the environment becomes available, the number

of cultivar parameters is likely to change and the definitions of some may also change.

2.4.2. Parameter definitions and values

Ecotype parameters are fully described in Table 3.2 in the scientific documentation. Parameter values

for a few ecotypes are proposed in Table 4.

Cultivar parameters are fully described in Table 3.3 in the scientific documentation. Parameter values

are proposed for a few real and hypothetical cultivars in Table 5. This is based on published reports

(Inman-Bamber, 1991; Singels & Bezuidenhout, 2002; Donaldson et al., 2003; Zhou et al., 2003 and

Zhou 2003) and unpublished data.

Often, too little information is available to choose an ecotype or determine values for the ecotype or

cultivar parameters. When this is the case, it is suggested that qualitative information about key traits

are used to estimate values. Table 6 proposes phenotypic trait categories for selected real and

hypothetical cultivars, while Table 7 suggests parameter values that will best emulate the required

phenotypic trait category.

33

Users who wish to create new cultivars within DSSAT without adequate quantitative information can

use the information in Tables 6 and 7 to guide them.

34

Table 4. Parameter values for the different ecotypes. Note: the greyed-out blocks indicate to the sugarcane model that the simpler ‘Canesim’ canopy option should be used. This is used when detailed leaf values are not available.

Parameter Category Description SC001 SC002 SC003 SC004 SC005 SC006

DELTTMAX Sucrose accumulation Max. change in sucrose content per unit change in stalk mass in the un-ripened section of the stalk (/t)

0.07

SWDF2AMP Sucrose accumulation Sucrose partitioning sensitivity to water stress parameter 0.5

CS_CNREDUC Canopy - CANESIM Canopy reduction due to water stress 0.3

CS_CNPERIOD Canopy - CANESIM Canopy water stress period (days) 21.

Tthalfa Canopy - CANESIM Half canopy thermal time adjustment for row width 125.

DPERdT Canopy - height Change in plant extension rate (mm/h) per unit change in temperature (oC) 0.176

EXTCFN Canopy - light extinction

Maximum canopy light extinction coefficient 0.84

EXTCFST Canopy - light extinction

Minimum canopy light extinction coefficient 0.58

LFNMXEXT Canopy - light extinction

Leaf number (including dead leaves still attached) at which maximum light extinction occurs

20.

AREAMX_CF(1) Canopy - leaves Cultivar parameter for quadratic equation defining maximum leaf area 0.

AREAMX_CF(2) Canopy - leaves Cultivar parameter for quadratic equation defining maximum leaf area 27.2

AREAMX_CF(3) Canopy - leaves Cultivar parameter for quadratic equation defining maximum leaf area -20.8

WIDCOR Canopy - leaves Parameter affecting the width of leaves 1.

WMAX_CF(1) Canopy - leaves Cultivar parameter for quadratic equation defining leaf width -0.0345

WMAX_CF(2) Canopy - leaves Cultivar parameter for quadratic equation defining leaf width 2.243

WMAX_CF(3) Canopy - leaves Cultivar parameter for quadratic equation defining leaf width 7.75

LMAX_CF(1) Canopy - leaves Cultivar parameter for quadratic equation defining max leaf length per leaf number -0.376

LMAX_CF(2) Canopy – leaves Cultivar parameter for quadratic equation defining max leaf length per leaf number 12.2

LMAX_CF(3) Canopy – leaves Cultivar parameter for quadratic equation defining max leaf length per leaf number 21.8

MAXLFLENGTH Canopy – leaves Absolute maximum leaf length 100

MAXLFWIDTH Canopy - leaves Absolute maximum leaf width 3.5

35

Table 4. (continued)

Parameter Category Description SC001 SC002 SC003 SC004 SC005 SC006

POPCF(1) Tiller population Stalk population coefficient, in ideal conditions (no stresses), as function of thermal time 1.826 1.4 1.6 2.0 2.2 1.826

POPCF(2) Tiller population Stalk population coefficient, in ideal conditions (no stresses), as function of thermal time -.002 -.0015 -.0018 -.002 -.002 -.002

POPDECAY Tiller population

Tiller senescence rate expressed as the fraction of tillers above the future mature tiller population (at a thermal time of 1600

oC.d), that senesce per unit thermal time

0.004

TTBASEEM Phenology Base temperature for emergence (oC) 10.

TTBASELFEX Phenology Base temperature for leaf phenology (oC) 10.

TTBASEPOP Phenology Base temperature for stalk phenology (oC) 16.

TBASEPER Phenology Base temperature for plant extension (oC) 10.057

LG_AMRANGE Lodging Range in aerial mass from the start to the end of lodging (t/ha) 30.

LG_GP_REDUC Lodging Reduction in gross photosynthesis due to full lodging, as a fraction (Singh, et al.) 0.28

LG_FI_REDUC Lodging Reduction in fractional interception by the canopy due to full lodging 0.1

36

Table 5. Proposed cultivar parameters for selected cultivars (actual and hypothetical).

Parameter Category Description

Nco

37

6

N31

N37

ZN

6

ZN

7

Q1

38

Q1

41

CG

1

CG

2

CG

3

Ecotype Determines tillering phenotype SC001 SC004 SC001 SC001 SC002 SC003 SC003 SC001 SC003 SC004

PARCEmax Biomass accumulation

Maximum (no stress) radiation conversion efficiency expressed as assimilate produced before respiration, per unit PAR (g/MJ).

9.9 9.9 9.9 9.9 9.9 9.9 9.9 9.9 9.9 9.9

APFMX Biomass partitioning

Maximum fraction of dry mass increments that can be allocated to aerial dry mass (t/t)

0.88 0.92 0.88 0.88 0.88 0.88 0.88 0.88 0.88 0.9

STKPFMAX Biomass partitioning

Fraction of daily aerial dry mass increments partitioned to stalk at high temperatures in a mature crop (t/t on a dry mass basis)

0.65 0.69 0.65 0.65 0.65 0.65 0.68 0.65 0.635 0.625

SUCA Sucrose accumulation

Sucrose partitioning parameter: Maximum sucrose contents in the base of stalk (t/t)

0.58 0.55 0.61 0.62 0.6 0.59 0.65 0.6 0.62 0.58

TBFT Sucrose accumulation

Sucrose partitioning: Temperature at which partitioning of unstressed stalk mass increments to sucrose is 50% of the maximum value

25 25 25 26 27 25 27 26 27 25

Tthalfo Canopy - CANESIM

Thermal time to half canopy (oC.d) 250 225 250 300 300 250 250 275 300 250

Tbase Canopy - CANESIM

Base temperature for canopy development (oC) 16 15 16 16 17 16 16 16 16 16

LFMAX Canopy - leaves Maximum number of green leaves a healthy, adequately-watered plant will have after it is old enough to lose some leaves.

12 12 12 12 12 12 12 12 12 12

MXLFAREA Canopy - leaves Max leaf area assigned to all leaves above leaf number MXLFARNO (cm

2)

360 390 480 420 420 360 360 500 600 600

MXLFARNO Canopy - leaves Leaf number above which leaf area is limited to MXLFAREA 14 14 14 16 18 14 14 16 17 17

PI1 Leaf phenology Phyllocron interval 1 (for leaf numbers below Pswitch,

oC.d (base

TTBASELFEX)) 69 69 90 100 100 69 69 110 110 90

PI2 Leaf phenology Phyllocron interval 2 (for leaf numbers above Pswitch,

oC.d (base

TTBASELFEX)) 169 169 170 200 200 169 169 200 200 170

PSWITCH Leaf phenology Leaf number at which the phyllocron changes 18 18 14 14 18 18 18 14 14 14

MAX_POP Tiller phenology Maximum tiller population (stalks/m2) 30 40 40 30 30 30 30 30 30 40

37

Table 5 (continued)

Parameter Category Description

Nco

37

6

N31

N37

ZN

6

ZN

7

Q1

38

Q1

41

CG

1

CG

2

CG

3

POPTT16 Tiller phenology Stalk population at/after 1600 degree days (/m2) 13.3 15.0 12.0 11.0 7.0 13.3 13.3 11.0 9.0 13.0

TTPLNTEM Phenology Thermal time to emergence for a plant crop (oC.d, base TTBASEEM) 428 428 428 428 428 428 428 428 475 375

TTRATNEM Phenology Thermal time to emergence for a ratoon crop (oC.d, base TTBASEEM) 203 203 203 203 203 203 203 203 250 150

CHUPIBASE Phenology Thermal time (baseTTBASEEM) from emergence to start of stalk growth

1050 790 1050 1150 1050 1050 1050 1050 1150 950

TT_POPGROWTH Phenology Thermal time to peak tiller population (oC.d, TTBASEPOP) 600 600 600 700 700 600 600 600 600 600

LG_AMBASE Lodging Aerial mass (fresh mass of stalks, leaves, and water attached to them) at which lodging starts; t/ha

220 220 220 220 220 220 220 220 220 220

38

Table 6. Subjective estimates (based on experimental data and expert opinion) of phenotypic trait values (categories) for selected

cultivars. Each trait is categorized into very (s)low (VL), (s)low (L), medium (M), high or rapid (H) or very high or rapid (VH).

Trait categories for hypothetical varieties are also suggested (CG1, CG2 and CG3).

Trait

Cultivars

CG

1

CG

2

CG

3

NC

o37

6

N31

N37

R570

ZN

6

ZN

7

N14

N19

N26

N12

N14

Q138

Q141

Primary tiller emergence

M L H M

Tiller population M L H H H H VL M VL M M L VH H M M

Leaf emergence M M H H M M VH H M H M H H H

Leaf size M H H L M H VH H M H H M L L L

Canopy development

M L H H H H H L H L L H H H

Biomass accumulation

M L H M H H M M M L L M M M M

Stalk growth M L H M H H M H M L L M M M M

Sucrose content M H L L VL H H M L H VH M L M H

Lodging tolerance M M M L

39

Table 7. Suggested values for key cultivar parameters for different phenotypic trait categories.

Trait category

Parameter Trait Description Very low

Low Med High Very high

PARCEmax Biomass accumulation Maximum (no stress) radiation conversion efficiency expressed as assimilate produced before respiration, per unit PAR. (g/MJ)

9.7 9.8 9.9 10.0 10.1

APFMX Biomass accumulation Maximum fraction of dry mass increments that can be allocated to aerial dry mass (t/t)

0.84 0.86 0.88 0.90 0.92

STKPFMAX Biomass partitioning Fraction of daily aerial dry mass increments partitioned to stalk at high temperatures in a mature crop (t/t on a dry mass basis)

0.62 0.635 0.65 0.675 0.68

SUCA Sucrose content Sucrose partitioning parameter: Maximum sucrose contents in the base of stalk (t/t) 0.56 0.58 0.6 0.62 0.64

TBFT Sucrose content Sucrose partitioning: Temperature at which partitioning of unstressed stalk mass increments to sucrose is 50% of the maximum value

24 25 26 27 28

Tthalfo Canopy development - CANESIM

Thermal time to half canopy (oC.d) 325 300 275 250 225

Tbase Canopy development - CANESIM

Base temperature for canopy development (oC) 18 17 16 15 14

LFMAX Canopy - leaves Maximum number of green leaves a healthy, adequately-watered plant will have after it is old enough to lose some leaves

10 11 12 13 13

MXLFAREA Leaf size Max leaf area assigned to all leaves above leaf number MXLFARNO (cm2) 300 400 500 600 700

MXLFARNO Leaf size Leaf number above which leaf area is limited to MXLFAREA 18 16 14 14 14

PI1 Leaf emergence Phyllocron interval 1 (for leaf numbers below Pswitch, oC.d (base TTBASELFEX)) 150 130 110 90 70

PI2 Leaf emergence Phyllocron interval 2 (for leaf numbers above Pswitch, oC.d (base TTBASELFEX)) 260 230 200 170 150

PSWITCH Leaf emergence Leaf number at which the phyllocron changes. 14 14 14 16 18

MAX_POP Tiller population Maximum tiller population (stalks/m2) 30 30 30 40 50

POPTT16 Tiller population Stalk population at/after 1600 degree days (/m2) 7 11 12 13 15

TTPLNTEM Primary tiller emergence Thermal time to emergence for a plant crop (oC.d, base TTBASEEM) 550 475 428 375 325

TTRATNEM Primary tiller emergence Thermal time to emergence for a ratoon crop (oC.d, base TTBASEEM) 250 225 200 150 100

TT_POPGROWTH Canopy development Thermal time to peak tiller population (oC.d, TTBASEPOP) 800 700 600 500 400

LG_AMBASE Lodging tolerance Aerial mass (fresh mass of stalks, leaves, and water attached to them) at which lodging starts; t/ha

220 230 240 250 270

40

2.4.3. Parameter calibration procedure

Cultivar information is split into two files – the cultivar file, and the ecotype file. The intention of the ecotype

file is to contain the same information for groups of cultivars; i.e. one ecotype definition serves many

cultivars. A set of cultivar parameters is intended to represent a cultivar for ALL field situations. It is

recommended that if any parameters are modified/adjusted, it should only be done on the basis of several

sets of measured experimental data across a variety of management regimes – different start/harvest dates,

locations, soils, etc.

If an ecotype definition is modified, bear in mind that these changes may affect other cultivar definitions. If in

doubt, always make a copy of an ecotype definition, rename/recode it, and modify this instead (instructions

provided below).

IMPORTANT NOTE CONCERNING GENETIC PARAMETERS: In the input files (cultivar and ecotype files),

all numbers are read as Fortran REAL numbers. Every number MUST therefore have a decimal point, even

integer values. This will be addressed in a later version of DSSAT-Canegro.

The following calibration procedure is proposed:

Ensure that soil, weather and management inputs are correct.

Compare simulated and observed values of the variables indicated below, in the order given below. If an

unacceptably large discrepancy exists, make an adjustment in the appropriate cultivar parameter(s). Two

approaches are possible here. Firstly appropriate experimental data can be analysed to derive the values

of parameters (explained below). The alternative, less preferred option is to qualitatively benchmark the

cultivar's phenotypic characteristics using Table 6 and then to select the appropriate value for key

parameters from Table 7.

2.4.4. Creating a cultivar definition

The following steps should be followed when attempting to define/determine new cultivar coefficients:

i. Copy the NCo376 cultivar line in the cultivar file to a new line in this file:

Open the cultivar file C:\DSSAT4\Genotype\SCCAN045.cul (select Notepad or Wordpad).

Select the NCo376 line – Edit [menu]copy

41

Scroll to the bottom of the screen and paste – Edit [menu] paste

ii. Increment the cultivar number (SC0000x) to one that has not yet been used:

2.1 Update cultivar code

3.1 Enter a new name

42

Type the new number into the first column; use spaces to ensure that the columns line up. THE

COLUMNS MUST LINE UP.

iii. Give the new cultivar definition a (unique) name:

Type this into the second column (again ensuring that the columns line up)

iv. Open the ecotype file and choose an ecotype whose parameters appear the closest match to those of

the cultivar whose parameters are attempting to be defined. Note the ecotype number (code).

Open the ecotype file - C:\DSSAT4\Genotype\SCCAN045.eco (using Notepad or Wordpad)

Search for an ecotype with desired characteristics – i.e. that match the characteristics of the actual

cultivar in the field (see discussion above)

Note the ecotype code/number

v. Back in the cultivar file, update the ecotype reference number in the cultivar line that has just been

added to the cultivar file.

Edit the ecotype reference in the new cultivar definition line to match the code of the ecotype just

identified.

vi. Test the simulation and examine tiller population (see below).

vii. If this is not appropriate, try changing the ecotype number in the cultivar definition to refer to a different

ecotype.

Ecotype codes

43

viii. If performance is still unsatisfactory, make a copy of the most suitable ecotype definition in the

ecotype file (copy the line and paste at the bottom), updating the name and number. Then update the

new cultivar definition to refer to this new ecotype definition – see below for more discussion of

creating new ecotype definitions.

ix. Experiment with different tiller population coefficients in the new ecotype definition (until the model

appears to simulate tiller population reasonably well).

x. Then examine leaf area index and canopy variables (see below), and adjust accordingly if necessary,

again working with new cultivar and/or ecotype definitions

xi. After this, address aboveground biomass parameters and performance (see below), again following a

similar procedure of adapting new cultivar and/or ecotype definitions

xii. Finally, follow a similar procedure for calibrating sucrose; details are presented below.

2.4.5. Creating new Ecotype definitions

If none of the existing ecotypes defined in the ecotype file are satisfactory, it will be necessary to define a

new ecotype definition. Follow a similar procedure to defining a new cultivar definition:

i. Open the ecotype file (C:\DSSAT4\Genotype\SCCAN045.ECO).

ii. Find the ecotype whose characteristics best match the desired characteristics of the new ecotype

definition.

iii. Copy this line and paste at the bottom of the file.

iv. Update the ecotype code number.

v. Edit the various parameter values accordingly; remember to always enter REAL values (i.e. with a

decimal point) and ensure that the columns line up.

vi. Save the file.

vii. Update the cultivar definition to use the new ecotype number in place of whatever was there before.

It is not highly recommended to change existing ecotype definitions – doing this might adversely affect many

cultivars.

2.4.6. Tiller population

Primary tiller emergence of plant or ratoon crop (TTPLNTEM and TTRATNEM): calculate the modelled

discrepancy in number of days and convert this to thermal time by using the average daily thermal time

observed during this period. Adjust the parameters by adding or subtracting this amount.

Tiller population curve parameters (POPCF1, POPCF2, POPTT16): select an ecotype by viewing the

parameters in the ecotype file (Table 4). The impact of these parameters on the tiller population curve is

shown in Figure 1.

44

Figure 1. Tiller population curves for one cultivar (NCo376) and five ecotypes (SC00*).

We have assumed that a strong correlation exists between rate of tillering and final stalk population.

The peak tiller population is regarded as a cultivar parameter and can be set in the cultivar field (MAX_POP

in Table 5). The curves shown in Figure. 1 would be capped at the value of MAX_POP. In the absence of

measured data we suggest that the value of MAX_POP does not 30 stalks/m2.

2.4.7. Canopy

The DSSAT-Canegro model has two canopy methods – the standard sophisticated canopy, where each leaf

on a stalks is modelled according to the parameters described below, and the simpler „Canesim‟ canopy,

which uses the Hill thermal time canopy model. The latter is invoked automatically if leaf parameters

(AREAMX_CF, WIDCOR, WMAX_CF) are left blank in the ecotype definition in the ecotype file. Stalk

population, although not used in the canopy calculations in such a case (when using the Canesim canopy),

is still calculated using Nco376 parameters and will be output in PlantGro.out.

2.4.8. Leaf area index

Leaf emergence parameters (PI1, PI2, PSWITCH): determine the values of these parameters by plotting

number of emerged leaves vs thermal time (base 10) and then calculating the inverse of the slope of two

linear regressions fitted to the data (see Figure. 2 for an example).

0

100

200

300

400

500

600

700

0 500 1000 1500 2000 2500

Thermal time (oC.d)

Till

er

popu

lation

(1

000/h

a)

SC002 (VL)

SC003 (L)

SC001 (M)

SC004 (H)

SC005 (VH)

NCo376

45

Figure 2. Number of emerged leaves as a function of thermal time (from Singels et al., 2005)

Leaf size parameters (MXLFAREA, MXLFARNO): determine the size (surface area) and the leaf number of

the biggest leaf from leaf size vs thermal time data.

Figure 3. An example of how to derive leaf size parameters from leaf size vs leaf

number data.

2.4.9. Aboveground biomass

Two options exist for adjusting the rate of aboveground biomass accumulation. Firstly, adjusting the

efficiency of converting PAR to biomass in the species file (PARCE) will alter the total amount of biomass

0

5

10

15

20

25

30

35

0 400 800 1200 1600 2000 2400 2800 3200 3600

Thermal time (oC.d)

Num

ber

of le

aves

per

shoot

NCo376 - Dec - R1

NCo376 - Jun - R1

Canegro

0

50

100

150

200

250

300

350

400

450

500

0 4 8 12 16 20 24

Leaf position from base of shoot

Le

af

are

a (

cm

2/le

af)

46

produced. Secondly, the partitioning of biomass to above-ground parts can be altered by adjusting the

partitioning fraction AFPMX. The relative response in biomass accumulation to adjustments will be similar to

the relative magnitude of the adjustments.

2.4.10. Stalk mass

Stalk partition fraction (STKPFMAX): determine the slope of the linear regression of stalk mass vs aerial dry

mass. An example is illustrated in Figure. 4.

Figure 4. An example of deriving the value of the stalk partition fraction from stalk

dry mass vs above ground biomass data.

Thermal time from primary tiller emergence to start of stalk elongation (CHUPIBASE): calculate thermal time

(base10) from the date of 50% primary tiller emergence to date of start of stalk elongation.

2.4.11. Sucrose mass

Sucrose parameters (SUCA, TBFT): select a parameter set for the different sucrose accumulation types from

Table 6. Increasing SUCA values will result in higher simulated sucrose contents (and vice versa), while

increasing TBFT results in a flatter (less response to temperature) and higher seasonal sucrose content

curve. Figure 5 shows typical seasonal sucrose curves for a low and high sucrose variety, under fully

irrigated conditions for Pongola, South Africa.

y = 0.6775x - 2.5938

R2 = 0.6541

0

5

10

15

20

25

30

35

40

45

50

0 10 20 30 40 50 60 70

Above ground biomass (t/ha)

Sta

lk d

ry m

ass (

t/h

a)

47

Figure 5. Actual and simulated stalk sucrose content (tons sucrose/tons stalk dry

matter) for a low and high sucrose content variety, fully irrigated at 12 months, at

Pongola, South Africa, (adapted from Singels, et al., 2005).

2.5. Experimental data

2.5.1. Overview of Experiment data files in DSSAT (for sugarcane)

One of the powerful features of DSSAT is the ability to store and analyse measured experimental data along

with simulated values. By comparing simulated and actual/measured values for a particular experiment (i.e. a

field experiment is conducted, then simulated in DSSAT), it is possible to:

Document experiments

Validate crop models

Visually/statistically assess model performance

Two kinds of measured/experimental data files are used in DSSAT: „T‟ files, which are „time course‟ files and

contain values for variables sampled throughout the season (e.g. daily soil water content, leaf area index,

etc.); and „A‟ files, which contain single end-of-season values for each variable (sucrose yield, stalk yield,

final soil water content, etc.) – these are either final values or average values over the course of the crop

(hence „A‟).

The FileT and FileA experimental data files are plain formatted-text files with a tabular structure. Column

headers define variables via data codes (which are explained in C:\DSSAT4\DATA.CDE), and the columns

of data contain measurements of these variables. Each row represents a sample for a particular treatment.

0.35

0.40

0.45

0.50

0.55

3 6 9 12

Harvest month

Su

cro

se c

on

ten

t

NCo376

N19

Canegro NCo376

Canegro H3

48

In a FileT (time course file), every record will feature a treatment number and the date that the sample was

taken:

@TRNO DATE STKD STKW SUCD

1 69210 24.70 109 10.14

1 69266 33.90 133 15.03

1 69315 34.70 130 15.34

1 70013 40.60 145 17.84

1 70069 136 15.23

1 70125 39.30 143 18.73

2 69266 17.80 80 6.56

2 69315 23.80 102 9.69

2 70013 31.50 132 13.60

2 70069 39.00 159 17.81

In file excerpt above, TRNO is the treatment number, DATE is the date in YRDOY format (e.g. 69210 is day

210 of 1969, which is 29 July 1969). STKD is stalk dry mass (t/ha), STKW is stalk fresh mass (t/ha), and

SUCD is sucrose dry mass (t/ha). The data definitions were looked up in DATA.CDE. The „@‟ sign identifies

the column headings.

In a FileA (Average / harvest data file), there is a single record for each treatment and no date:

@TRNO SUCH TRSH GLAI STKH CHTA L#SM HIAM AELH SDAH

1 16.80 10.90 2.6 34.70 1.792 7.8 .482 52.80 13.2

2 13.00 8.200 2.6 26.70 1.685 9.2 .487 41.00 13.2

3 14.70 10.80 2.8 33.40 2.083 9.8 .440 50.40 14.5

4 14.70 10.10 3.7 34.70 2.209 10.4 .422 52.80 13.4

5 19.20 13.30 3.7 42.50 2.360 8.9 .452 62.80 15.5

6 19.90 15.30 2.1 38.30 2.223 6.7 .521 58.90 14.9

7 16.50 12.10 1.7 30.50 1.935 7.4 .540 46.30 12.9

8 14.10 12.80 1.3 27.90 2.216 9.2 .506 45.10 10.8

In the excerpt above, a single set of measurements is listed for each treatment (TRNO). SUCH is sucrose

mass at harvest (t/ha), TRSH is trash mass at harvest (t/ha), GLAI is final leaf area index (m2/m

2), STKH is

stalk dry mass at harvest (t/ha), CHTA is canopy height at harvest, etc. (the data definitions were looked up

in DATA.CDE).

49

FileA data appear in the Overview.OUT file generated by the DSSAT model when it runs. Here is an

example:

*MAIN GROWTH AND DEVELOPMENT VARIABLES

@ VARIABLE SIMULATED MEASURED

-------- --------- --------

Sucrose dry mass (t/ha) at harvest 12.00 16.80

Aerial dry biomass (t/ha) at harvest 44.89 52.80

Stalk (millable) dry mass, t/ha) at harvest 24.88 34.70

Trash (residue) dry mass (t/ha) at harvest 7.62 10.90

Green leaf area index (m2/m2) 2.16 2.6

Leaf area index, maximum 3.76 -99

Canopy height (m) 2.62 1.792

Harvest index at maturity 0.48 .482

Leaf number per stem at maturity 9.67 7.8

FileA data also appear in EVALUATE.OUT.

FileT (time course) data are used primarily in the graphing program, Gbuild. Measured data points, being

independent samples, are plotted as single points on the graphs. Simulated data, being determined on a

continuous daily basis, are plotted as lines.

2.5.2. Entering FileT and FileA files

Data are collected in various ways from the field. Electronic instruments generally either download to text

files or spreadsheets. Sensors are also frequently connected to dataloggers, which are downloaded to text

files. Measurements are also taken by hand and captured into databases and spreadsheets. Two avenues

are available to the DSSAT user for entering the data into DSSAT:

Using the DSSAT ATCreate program

Using a text editor (e.g. Notepad) to create the files manually

The ATCreate program is slightly unstable and the experience of the authors suggests that users take

particular care when using this software. Editing the files manually can be an arduous task, and ATCreate is

the recommended approach. However , it is recommended that:

Users make use of Excel to perform all data processing such that ATCreate has simply to convert from

an Excel / CSV document into FileT or FileA format.

Any subsequent editing is performed manually with a text editor.

The process of creating T and A files is exactly the same for sugarcane as any other crop in DSSAT, so the

reader is directed to the general DSSAT v4 and v4.5 documentation in this regard.

50

3. Simulation settings

When the crop management / experiment file (FileX) is set up, the user describes the experimental crop

management setup such that the crop model has enough information to run a simulation. This crop

management setup includes planting details, choice of cultivars, when to harvest, and so on. As a whole,

these can be considered „management‟ configuration.

It is also necessary, however, to provide some MODEL configuration – guidelines to the crop model itself as

to how it should go about running the simulation. These guidelines take the form of, for example, how to

schedule irrigation; how reference evaporation should be calculated; on what date should the simulation

start; and so on. Complication is introduced in some cases where these guidelines correspond with

management settings – if a harvest factor level (i.e. harvest date) is defined, for example, the model needs to

be explicitly instructed to end the simulation on the specified harvest date; otherwise, DSSAT will wait for the

crop itself to „mature‟ and end the simulation itself, which is the default DSSAT CSM behaviour.

The user is invited to explore the options presented under the Simulation Options menu in Xbuild. Simulation

Options are treated as a factor, and a set of Simulation Options is associated as a factor level with each

treatment. In this way, irrigation regimes can be compared in different treatments, and so on.

3.1. Specific Sugarcane simulation options

Some simulation options are particularly relevant to sugarcane. It is important that these are set. Failure to

do so will either result in the model crashing or the model simulating poorly:

Harvest: In the Simulation Options [menu] Management [button] Harvest [button] section, the

simulation MUST be instructed to harvest „On reported date‟ OR „Days after planting‟. This choice

MUST correspond with a harvest crop management factor level setting (Management menu

„Harvests‟) – see section 2.3, step 6.

Planting: In the Simulation Options [menu] Management [button] Planting [button] section, the

simulation MUST be instructed to plant „On reported date‟. This choice MUST correspond with a

planting crop management factor level setting (Management menu „Planting‟) – see section 2.3,

step 5.

Reference evaporation: If TDEW (dewpoint temperature) or relative humidity and windspeed are

available in the weather data used by this simulation, the FAO-56 reference evaporation method

should be chosen. Set this by clicking Simulation Options [menu] Methods [button]

Evapotranspiration [drop down].

Plant/ratoon crops: if a plant crop is to be simulated, the correct choice on the planting screen is „Dry

seed‟. „Ratoon‟ can be chosen on this list as well for a ratoon crop.

3.2. Irrigation

Irrigation details are generally set in the Simulation Options section of Xbuild. The only exception to this is if

a record of daily irrigation amounts was kept. These must be entered in the Management [menu]Irrigation

screen, but even then the simulation option for irrigation must still be set. In this specific case the irrigation

management under the Simulation Options menu should be set to „On reported dates‟.

Irrigation will usually be automatic. If extremely generous irrigation is desired, set the management depth to

a small value and the threshold percentage to a high value, in the irrigation management section under

Simulation Options. The top of the soil profile dries out faster, so this will result in more frequent irrigations.

51

Please see the Management section (2.3, step 8) for an example of how simulation options can be used.

4. Running the model and viewing outputs

4.1. Running the sugarcane model

The sugarcane model is now part of the DSSAT Cropping System Model, so behaves like any other crop in

terms of the user experience.

i. In DSSAT, navigate to the sugarcane section using the navigation tree on the mid-left side of the

screen.

ii. Click the Refresh button to ensure that experiment lists are updated.

iii. Locate the experiment (by code or description); click on the checkbox next to it.

iv. The treatments belonging to this experiment will be listed in the pane below. Ensure that the

treatments that need to be run have their checkboxes ticked.

v. Click on the Run button at the above-left of the screen:

vi. You will then get this screen:

52

vii. Click on the „Run Model‟ button to run the simulation; a DOS-prompt screen will briefly appear:

viii. When the DOS screen disappears, the simulation is complete.

4.2. Viewing and graphing model output

In order to visulaise model output, the following steps are required:

i. Back on the Run screen, click on the „Analysis‟ tab; this displays a list of .OUT files in the DSSAT

sugarcane directory.

53

ii. Now click on the checkbox next to „PlantGro.OUT‟:

iii. Clicking on the „View‟ button will load up any checked files in Notepad. If the „Plot‟ button is pressed,

the Gbuild graphing program opens with the checked files loaded (although only one file‟s variables

can be graphed at a time).

iv. The Gbuild screen presents a list of variables in the loaded file on the lefthand side of the screen, and

a list of treatments on the right. Click on the checkboxes next to the variables that need to be graphed,

mark the checkboxes next to the treatments for which these variables need to be graphed.

54

v. The figure below shows a sample of an output graph:

55

vi. Clicking on the Back button brings the user back to the variable/treatment selection screen. If

experimental data are available for this experiment and the currently-graphed treatments, these data

points will be graphed in the same colour as the corresponding model-simulated series line. The

Statistic button will be enabled, and if this is pressed, Gbuild calculates statistics such as RMSE,

mean, d-stat, etc. for each treatment.

4.3. General discussion of model outputs

The sugarcane model in DSSAT (the Canegro CSM module) produces sugarcane-specific output. This

output is found in the Plantgro.OUT file, primarily; Plantgro.OUT contains all of the plant information – yields,

root mass, water stress, canopy cover, and so on. Other outputs are put in OVERVIEW.OUT and INFO.OUT.

DSSAT also generates ET.OUT (output variables associated with evapotranspiration), WATBAL.OUT

(variables associated with soil water content), Weather.OUT (daily weather values), and various others.

In any output file, the definition of a variable output code can be looked up in C:\DSSAT4\DATA.CDE.

All output from a sugarcane simulation is stored in C:\DSSAT4\sugarcane. Other sugarcane-specific files are

also stored here – e.g. the sugarcane FileX files. Soil and weather files, on account of being able to be used

for other crops too, are stored in C:\DSSAT4\Soil and C:\DSSAT4\Weather respectively. If a sequence of

crops is simulated (whether not sugarcane is in this sequence), output will always go into

C:\DSSAT4\Sequence. Cultivar, ecotype and species files are stored in C:\DSSAT4\Genotype.

The DSSAT CSM model (which contains the sugarcane module) can also be run from the command line.

When DSSAT is run from the user interface program, a file called d4batch.dv4 is created. This contains a list

of runs that DSSAT must execute. The model could then be run with the following commands in DOS:

c:

cd \DSSAT4\sugarcane

c:\DSSAT4\DSCSM045.exe b d4batch.dv4

This is running the CSM executable called DSCSM045.exe, which is stored in C:\DSSAT4. If all requisite

files – exe, soil, weather, cultivar, FileX, dv4, etc – are put in an arbitrary directory, the model can run from

that directory, in the form:

C:

cd \arbitrary_directory\

DSCSM045.exe b d4batch.dv4

It can be extremely useful to run the model in this way. The design of DSSAT is such that treatments from

many different experiments (and crops) can run from a single batch (dv4) file. All output goes into single files

as well, generally making analysis much easier.

5. Simulating a sequence of plant and ratoon crops and a fallow period.

5.1. Introduction

This section describes how DSSAT can be used to simulate a sugarcane replant cycle. This will involve

planting a sugarcane crop, harvesting, allowing it to ratoon/harvest three times, leaving it for a three-month

fallow, replanting, and running for another three ratoons. An example is used to illustrate the steps required

to run this simulation.

56

It is assumed that the user is familiar with setting up DSSAT simulations. Only the parts that differ for a

plant/ratoon sequence will be discussed.

5.2. Scenario

Weather data: Tongaat automatic weather station, KZN, South Africa

Soil: Arcadia (Mount Edgecombe)

Crop: Sugarcane, fallow

Cultivar: NCo376

Dates: See Table 8.

Table 8. Dates for simulating a sugarcane cropping sequence

Event Start Harvest

Plant 1 December 2000 1 December 2001

Ratoon 1 2 December 2001 2 December 2002

Ratoon 2 3 December 2002 3 December 2003

Fallow 4 December 2003 4 March 2004

Plant 5 March 2004 5 March 2005

5.3. Method

5.3.1. Create a FileX

Use the Xbuild program to create a new FileX.

5.3.2. Experiment details

This is a crop sequence (the fallow is treated as a crop), so the ‘Sequence’ option is to be chosen in XBuild.

5.3.3. Crops

Because this is a sequence, each crop in the sequence must be added. Choose „Sugarcane‟, „NCo376‟, for

Sequence Level 1, and „Fallow‟, „fallow‟ for fallow period.

5.3.4. Planting

Create a planting „factor level‟ called „Plant cane (1)‟, representing the first sugarcane plant crop. Enter start

date from table above (1 December 2000). Planting method is set as „dry seed‟. Click „Add‟ to add another

planting factor level. This is the ratoon cane; all settings are the same, except the date (2 December 2001),

and the planting method (ratoon). Add in all ratoon plant dates and plant plant dates as planting factor levels.

5.3.5. Harvests

Click on the „Management‟ menu, and select „Harvest‟. Enter each harvest date, and a harvest „level‟ for

each one. Add in harvest dates for all plants and ratoons.

57

5.3.6. Treatments

Create a new treatment rotation for each plant/harvest combination, i.e. one for the first planting and harvest,

one for the first ratoon and harvest, etc. In this case, four treatment rotations are set up.

5.3.7. Simulation options

Set start sim date, replications (1), years (4). Set „on reported dates‟ for plant and harvest. No applications of

irrigation or fertiliser. Options: simulate water, not Nitrogen. Output: 1 day frequency, growth, water. Four-

year run (not 5), because this simulation fits within four years of weather data (and the weather data does

not span from the first planting date to first planting date + 5 years). Weather data must be „measured‟, not

generated.

Notes: This may seem obvious, but make sure the weather data set spans the entire duration of the

simulation. It is easy to mistakenly set DSSAT to „overrun‟, i.e. attempt to simulate beyond the last available

weather date.

6. Acknowledgements

The valuable contributions to team members of the ICSM project to incorporate the Canegro model into

DSSAT4.5 are acknowledged. Maurits van den Berg (SASRI), Graham Kingston (BSES), Attachai Jintrawet

(Chang Mai University), Simbarashe Chinorumbe (ZSAES), Jim Shine (SCGC) and Michiel Smit (SASRI)

have assisted in improving the format and organization of input and output data; in highlighting problem

areas and bugs; and testing the model for a wide range of conditions and cultivars. The assistance and

guidance of members of the DSSAT team, namely Cheryl Porter and Jim Jones (UF), are also gratefully

acknowledged.

The financial assistance of the ICSM via its members for this project is gratefully acknowledged.

7. References

Donaldson, R.A., Redshaw, K.A. and Singels, A., 2003. In search of parameters to model cultivar specific

canopy development. Proc S Afr Sug Technol Ass 77:605-609.

Inman-Bamber, N.G., 1991. A growth model for sugarcane based on a simple carbon balance and the

CERES-Maize water balance. S. Afr. J. Plant Soil, 8: 93-99.

58

Singels, A. & Bezuidenhout, C.N., 2002. A new method of simulating dry matter partitioning in the Canegro

sugarcane model. Field Crops Research 78: 151–164.

Singels, A., Jones, MR. and van den Berg, M., 2008. DSSAT v4.5 - Canegro Sugarcane Plant Module:

Scientific documentation. ICSM report.

Singels, A., Smit, M.A., Redshaw, K.A. & Donaldson, R.A., 2005. The effect of crop start date, crop class and

cultivar on sugarcane canopy development and radiation interception. Field Crops Res. 92: 249-260.

Zhou, M.M., 2003. Modelling variety differences in canopy growth and development of sugarcane

(Saccharum officinarum L.). M.Sc. dissertation. Faculty of Science and Agriculture, University of KwaZulu-

Natal, Pietermaritzburg, South Africa.

Zhou, M.M., Singels, A., Savage, M.J., 2003. Physiological parameters for modelling differences in canopy

development between sugarcane cultivars. Proc. S. Afr. Sug. Technol. Ass. 77:610-621.

Documents

DSSAT